Proving Convergence of Averages of a Convergent Sequence

Click For Summary

Homework Help Overview

The problem involves proving that the sequence of averages of a convergent sequence also converges to the same limit. The subject area is real analysis, specifically focusing on sequences and limits.

Discussion Character

  • Exploratory, Mathematical reasoning, Assumption checking

Approaches and Questions Raised

  • Participants discuss how to show that the average sequence converges by estimating the difference between the averages and the limit of the original sequence. Some suggest using properties of limits and inequalities to establish bounds.

Discussion Status

Multiple approaches are being explored, including bounding the averages and breaking down the terms involved. Participants are providing insights and suggestions without reaching a consensus on a specific method.

Contextual Notes

There is a mention of needing to consider terms before a certain index N, and the discussion includes the challenge of handling finite sums versus infinite limits.

R.P.F.
Messages
210
Reaction score
0

Homework Statement



Show that if (x_n) is a convergent sequence, then the sequence given by the averages

y_n = (x_1+ x_2 +...+ x_n)/n

also converges to the same limit.

Homework Equations





The Attempt at a Solution



I think I need to show that for n >= N, |y_n - lim(x_n)| < e for an arbitrary e > 0. I don't know where t start because y_n involves terms before N.
Any help is appreciated! :)
 
Physics news on Phys.org
I'm not sure that this really works, but here's something you might try...

If you call x the limit of xn, you can estimate
[tex]|y_n - x| = \left| \frac{x_1}{n} + \frac{x_2}{n} + \cdots + \frac{x_n}{n} - x \right| \le \frac{1}{n} \left( |x_1 - x| + |x_2 - x| + \cdots |x_n - x| \right)[/tex]

At least then you'll have written everything in terms of |xi - x| which occurs in the limit definition of xn.

Another thought that occurred to me while writing this message is that you could take the maximum of the (finitely many) xi, and write
[tex](x_1 + x_2 + \cdots + x_n) / n \le n x_i / n = x_i[/tex]
for some i = 1, 2, ..., n.
 
A traditional (brute force in a sense) approach: Since [tex]x_n \to L[/tex], for every [tex]\epsilon>0[/tex] there exists [tex]n_0\in \mathbb{N}[/tex] such that for all [tex]n>n_0[/tex] we have [tex]L-\epsilon < x_n < L + \epsilon[/tex]

Then for [tex]n>n_0[/tex] we write:

[tex]\frac{ x_1 + x_2 + ... + x_n}{n} = \frac{ x_1 + x_2 + ... +x_{n_0}}{n} + \frac{x_{n_0 +1} + x_{n_0 +2} + ... + x_n}{n}[/tex]

For the first term, the numerator adds to some finite number, so we can use the Archimedian property to bound that, and for the second term we can use the inequalities above to bound that.
 
Thank you so much, guys!
May I ask how to use LaTeX on PF? Thanks!
 
If you click on the LaTex you can see the code that we typed to display what you see. Also there's a reference to a brief guide underneath the code. You'll pick it up by clicking on other peoples images and learning the basic structure of the code.
 
Thank you, Gib. I know how to write LaTeX code. I'm just wondering how to post it onto PF. Do I just add [tex]and [\tex]?[/tex]
 
Yup just those before and after the code. Or for inline, [ itex] and [ /itex]
 

Similar threads

  • · Replies 8 ·
Replies
8
Views
3K
Replies
8
Views
3K
Replies
11
Views
2K
  • · Replies 6 ·
Replies
6
Views
2K
  • · Replies 7 ·
Replies
7
Views
2K
  • · Replies 13 ·
Replies
13
Views
2K
Replies
4
Views
2K
  • · Replies 2 ·
Replies
2
Views
2K
Replies
3
Views
2K
  • · Replies 11 ·
Replies
11
Views
3K